[ANNOUNCE] 3.0.6-rt16

From: Thomas Gleixner
Date: Tue Oct 04 2011 - 16:20:03 EST

Dear RT Folks,

I'm pleased to announce the 3.0.6-rt16 release.

Changes from 3.0.4-rt14 to 3.0.4-rt15

This is an intermediate version to get a rt only delta.
Please use 3.0.4-rt16!

* Drop RCU_bh hack (Thanks to Paul McKenney for review and explanation!)

* Hotplug fix for softirq (Peter Zijlstra) - should fix cpu
online/offline and poweroff/suspend problems

* migrate_disable() performance optimizations (Steven Rostedt, Peter Zijlstra)

* workqueue sanitizing (Peter Zijlstra)

* Reenable adaptive rtmutex spinning

* ARM split PTL fix (Frank Rowand)

* MIPS IRQ_NO_THREAD annotations (Venkat Subbiah)

* x86_64 irq vs. idle_exit fixes (Frederic Weisbecker)

* tracing fixes (Steven Rostedt)

* Forced sysrq printing (Frank Rowand)

* HPET MSI disable for Lenovo W510

* Raw spinlock annotations (Frank Rowand)

* !RT compile fix (John Kacur)

* Rtmutex debug RCU fix

* Some cherry picked mainline fixes

Delta patch against 3.0.4-rt14

https://tglx.de/~tglx/rt/older/patch-3.0.4-rt14-rt15.patch.gz

also appended below.


Changes from 3.0.4-rt15 to 3.0.6-rt16

* Update to 3.0.6 (Dropped a few patches which made it into 3.0.6)


Patch against 3.0.6 can be found here:

https://tglx.de/~tglx/rt/patch-3.0.6-rt16.patch.gz


The split quilt queue is available at:

https://tglx.de/~tglx/rt/patches-3.0.6-rt16.tar.gz

Enjoy,

tglx

--------------->
Index: linux-2.6/arch/arm/kernel/process.c
===================================================================
--- linux-2.6.orig/arch/arm/kernel/process.c
+++ linux-2.6/arch/arm/kernel/process.c
@@ -484,6 +484,31 @@ unsigned long arch_randomize_brk(struct
}

#ifdef CONFIG_MMU
+
+/*
+ * CONFIG_SPLIT_PTLOCK_CPUS results in a page->ptl lock. If the lock is not
+ * initialized by pgtable_page_ctor() then a coredump of the vector page will
+ * fail.
+ */
+static int __init vectors_user_mapping_init_page(void)
+{
+ struct page *page;
+ unsigned long addr = 0xffff0000;
+ pgd_t *pgd;
+ pud_t *pud;
+ pmd_t *pmd;
+
+ pgd = pgd_offset_k(addr);
+ pud = pud_offset(pgd, addr);
+ pmd = pmd_offset(pud, addr);
+ page = pmd_page(*(pmd));
+
+ pgtable_page_ctor(page);
+
+ return 0;
+}
+late_initcall(vectors_user_mapping_init_page);
+
/*
* The vectors page is always readable from user space for the
* atomic helpers and the signal restart code. Let's declare a mapping
Index: linux-2.6/arch/arm/plat-mxc/include/mach/iomux-v3.h
===================================================================
--- linux-2.6.orig/arch/arm/plat-mxc/include/mach/iomux-v3.h
+++ linux-2.6/arch/arm/plat-mxc/include/mach/iomux-v3.h
@@ -66,6 +66,7 @@ typedef u64 iomux_v3_cfg_t;
#define MUX_MODE_MASK ((iomux_v3_cfg_t)0x1f << MUX_MODE_SHIFT)
#define MUX_PAD_CTRL_SHIFT 41
#define MUX_PAD_CTRL_MASK ((iomux_v3_cfg_t)0x1ffff << MUX_PAD_CTRL_SHIFT)
+#define NO_PAD_CTRL ((iomux_v3_cfg_t)1 << (MUX_PAD_CTRL_SHIFT + 16))
#define MUX_SEL_INPUT_SHIFT 58
#define MUX_SEL_INPUT_MASK ((iomux_v3_cfg_t)0xf << MUX_SEL_INPUT_SHIFT)

@@ -84,7 +85,6 @@ typedef u64 iomux_v3_cfg_t;
* Use to set PAD control
*/

-#define NO_PAD_CTRL (1 << 16)
#define PAD_CTL_DVS (1 << 13)
#define PAD_CTL_HYS (1 << 8)

Index: linux-2.6/arch/x86/kernel/apic/io_apic.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/apic/io_apic.c
+++ linux-2.6/arch/x86/kernel/apic/io_apic.c
@@ -2275,8 +2275,8 @@ asmlinkage void smp_irq_move_cleanup_int
unsigned vector, me;

ack_APIC_irq();
- exit_idle();
irq_enter();
+ exit_idle();

me = smp_processor_id();
for (vector = FIRST_EXTERNAL_VECTOR; vector < NR_VECTORS; vector++) {
Index: linux-2.6/arch/x86/kernel/cpu/mcheck/mce.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/cpu/mcheck/mce.c
+++ linux-2.6/arch/x86/kernel/cpu/mcheck/mce.c
@@ -471,8 +471,8 @@ static inline void mce_get_rip(struct mc
asmlinkage void smp_mce_self_interrupt(struct pt_regs *regs)
{
ack_APIC_irq();
- exit_idle();
irq_enter();
+ exit_idle();
mce_notify_irq();
mce_schedule_work();
irq_exit();
Index: linux-2.6/include/linux/cpu.h
===================================================================
--- linux-2.6.orig/include/linux/cpu.h
+++ linux-2.6/include/linux/cpu.h
@@ -60,14 +60,16 @@ enum {
*/
CPU_PRI_SCHED_ACTIVE = INT_MAX,
CPU_PRI_CPUSET_ACTIVE = INT_MAX - 1,
- CPU_PRI_SCHED_INACTIVE = INT_MIN + 1,
- CPU_PRI_CPUSET_INACTIVE = INT_MIN,

/* migration should happen before other stuff but after perf */
- CPU_PRI_PERF = 20,
- CPU_PRI_MIGRATION = 10,
- /* prepare workqueues for other notifiers */
- CPU_PRI_WORKQUEUE = 5,
+ CPU_PRI_PERF = 20,
+ CPU_PRI_MIGRATION = 10,
+ CPU_PRI_WORKQUEUE_ACTIVE = 5, /* prepare workqueues for others */
+ CPU_PRI_NORMAL = 0,
+ CPU_PRI_WORKQUEUE_INACTIVE = -5, /* flush workqueues after others */
+
+ CPU_PRI_SCHED_INACTIVE = INT_MIN + 1,
+ CPU_PRI_CPUSET_INACTIVE = INT_MIN,
};

#ifdef CONFIG_SMP
Index: linux-2.6/include/linux/rcupdate.h
===================================================================
--- linux-2.6.orig/include/linux/rcupdate.h
+++ linux-2.6/include/linux/rcupdate.h
@@ -78,13 +78,7 @@ struct rcu_head {
extern void call_rcu_sched(struct rcu_head *head,
void (*func)(struct rcu_head *rcu));
extern void synchronize_sched(void);
-
-#ifdef CONFIG_PREEMPT_RT_FULL
-# define rcu_barrier_bh rcu_barrier
-#else
extern void rcu_barrier_bh(void);
-#endif
-
extern void rcu_barrier_sched(void);

static inline void __rcu_read_lock_bh(void)
@@ -144,13 +138,7 @@ static inline int rcu_preempt_depth(void

/* Internal to kernel */
extern void rcu_sched_qs(int cpu);
-
-#ifndef CONFIG_PREEMPT_RT_FULL
extern void rcu_bh_qs(int cpu);
-#else
-static inline void rcu_bh_qs(int cpu) { }
-#endif
-
extern void rcu_check_callbacks(int cpu, int user);
struct notifier_block;

@@ -241,14 +229,7 @@ static inline int rcu_read_lock_held(voi
* rcu_read_lock_bh_held() is defined out of line to avoid #include-file
* hell.
*/
-#ifdef CONFIG_PREEMPT_RT_FULL
-static inline int rcu_read_lock_bh_held(void)
-{
- return rcu_read_lock_held();
-}
-#else
extern int rcu_read_lock_bh_held(void);
-#endif

/**
* rcu_read_lock_sched_held() - might we be in RCU-sched read-side critical section?
@@ -657,13 +638,8 @@ static inline void rcu_read_unlock(void)
static inline void rcu_read_lock_bh(void)
{
__rcu_read_lock_bh();
-
-#ifdef CONFIG_PREEMPT_RT_FULL
- rcu_read_lock();
-#else
__acquire(RCU_BH);
rcu_read_acquire_bh();
-#endif
}

/*
@@ -673,12 +649,8 @@ static inline void rcu_read_lock_bh(void
*/
static inline void rcu_read_unlock_bh(void)
{
-#ifdef CONFIG_PREEMPT_RT_FULL
- rcu_read_unlock();
-#else
rcu_read_release_bh();
__release(RCU_BH);
-#endif
__rcu_read_unlock_bh();
}

@@ -785,9 +757,6 @@ extern void call_rcu(struct rcu_head *he

#endif /* #else #ifdef CONFIG_PREEMPT_RCU */

-#ifdef CONFIG_PREEMPT_RT_FULL
-#define call_rcu_bh call_rcu
-#else
/**
* call_rcu_bh() - Queue an RCU for invocation after a quicker grace period.
* @head: structure to be used for queueing the RCU updates.
@@ -808,7 +777,6 @@ extern void call_rcu(struct rcu_head *he
*/
extern void call_rcu_bh(struct rcu_head *head,
void (*func)(struct rcu_head *head));
-#endif

/*
* debug_rcu_head_queue()/debug_rcu_head_unqueue() are used internally
Index: linux-2.6/include/linux/rcutree.h
===================================================================
--- linux-2.6.orig/include/linux/rcutree.h
+++ linux-2.6/include/linux/rcutree.h
@@ -57,11 +57,7 @@ static inline void exit_rcu(void)

#endif /* #else #ifdef CONFIG_TREE_PREEMPT_RCU */

-#ifndef CONFIG_PREEMPT_RT_FULL
extern void synchronize_rcu_bh(void);
-#else
-# define synchronize_rcu_bh() synchronize_rcu()
-#endif
extern void synchronize_sched_expedited(void);
extern void synchronize_rcu_expedited(void);

@@ -75,18 +71,12 @@ extern void rcu_barrier(void);
extern unsigned long rcutorture_testseq;
extern unsigned long rcutorture_vernum;
extern long rcu_batches_completed(void);
+extern long rcu_batches_completed_bh(void);
extern long rcu_batches_completed_sched(void);

extern void rcu_force_quiescent_state(void);
-extern void rcu_sched_force_quiescent_state(void);
-
-#ifndef CONFIG_PREEMPT_RT_FULL
extern void rcu_bh_force_quiescent_state(void);
-extern long rcu_batches_completed_bh(void);
-#else
-# define rcu_bh_force_quiescent_state rcu_force_quiescent_state
-# define rcu_batches_completed_bh rcu_batches_completed
-#endif
+extern void rcu_sched_force_quiescent_state(void);

/* A context switch is a grace period for RCU-sched and RCU-bh. */
static inline int rcu_blocking_is_gp(void)
Index: linux-2.6/include/linux/rwlock_types.h
===================================================================
--- linux-2.6.orig/include/linux/rwlock_types.h
+++ linux-2.6/include/linux/rwlock_types.h
@@ -47,6 +47,7 @@ typedef struct {
RW_DEP_MAP_INIT(lockname) }
#endif

-#define DEFINE_RWLOCK(x) rwlock_t x = __RW_LOCK_UNLOCKED(x)
+#define DEFINE_RWLOCK(name) \
+ rwlock_t name __cacheline_aligned_in_smp = __RW_LOCK_UNLOCKED(name)

#endif /* __LINUX_RWLOCK_TYPES_H */
Index: linux-2.6/kernel/fork.c
===================================================================
--- linux-2.6.orig/kernel/fork.c
+++ linux-2.6/kernel/fork.c
@@ -87,7 +87,7 @@ int max_threads; /* tunable limit on nr

DEFINE_PER_CPU(unsigned long, process_counts) = 0;

-__cacheline_aligned DEFINE_RWLOCK(tasklist_lock); /* outer */
+DEFINE_RWLOCK(tasklist_lock); /* outer */

#ifdef CONFIG_PROVE_RCU
int lockdep_tasklist_lock_is_held(void)
Index: linux-2.6/kernel/printk.c
===================================================================
--- linux-2.6.orig/kernel/printk.c
+++ linux-2.6/kernel/printk.c
@@ -21,6 +21,7 @@
#include <linux/tty.h>
#include <linux/tty_driver.h>
#include <linux/console.h>
+#include <linux/sysrq.h>
#include <linux/init.h>
#include <linux/jiffies.h>
#include <linux/nmi.h>
@@ -831,8 +832,8 @@ static int console_trylock_for_printk(un
__releases(&logbuf_lock)
{
#ifdef CONFIG_PREEMPT_RT_FULL
- int lock = !early_boot_irqs_disabled && !irqs_disabled_flags(flags) &&
- !preempt_count();
+ int lock = (!early_boot_irqs_disabled && !irqs_disabled_flags(flags) &&
+ !preempt_count()) || sysrq_in_progress;
#else
int lock = 1;
#endif
Index: linux-2.6/kernel/rcupdate.c
===================================================================
--- linux-2.6.orig/kernel/rcupdate.c
+++ linux-2.6/kernel/rcupdate.c
@@ -72,7 +72,6 @@ int debug_lockdep_rcu_enabled(void)
}
EXPORT_SYMBOL_GPL(debug_lockdep_rcu_enabled);

-#ifndef CONFIG_PREEMPT_RT_FULL
/**
* rcu_read_lock_bh_held() - might we be in RCU-bh read-side critical section?
*
@@ -92,7 +91,6 @@ int rcu_read_lock_bh_held(void)
return in_softirq() || irqs_disabled();
}
EXPORT_SYMBOL_GPL(rcu_read_lock_bh_held);
-#endif

#endif /* #ifdef CONFIG_DEBUG_LOCK_ALLOC */

Index: linux-2.6/kernel/rcutree.c
===================================================================
--- linux-2.6.orig/kernel/rcutree.c
+++ linux-2.6/kernel/rcutree.c
@@ -166,7 +166,6 @@ void rcu_sched_qs(int cpu)
rdp->passed_quiesc = 1;
}

-#ifndef CONFIG_PREEMPT_RT_FULL
void rcu_bh_qs(int cpu)
{
struct rcu_data *rdp = &per_cpu(rcu_bh_data, cpu);
@@ -175,7 +174,6 @@ void rcu_bh_qs(int cpu)
barrier();
rdp->passed_quiesc = 1;
}
-#endif

/*
* Note a context switch. This is a quiescent state for RCU-sched,
@@ -218,7 +216,6 @@ long rcu_batches_completed_sched(void)
}
EXPORT_SYMBOL_GPL(rcu_batches_completed_sched);

-#ifndef CONFIG_PREEMPT_RT_FULL
/*
* Return the number of RCU BH batches processed thus far for debug & stats.
*/
@@ -236,7 +233,6 @@ void rcu_bh_force_quiescent_state(void)
force_quiescent_state(&rcu_bh_state, 0);
}
EXPORT_SYMBOL_GPL(rcu_bh_force_quiescent_state);
-#endif

/*
* Record the number of times rcutorture tests have been initiated and
@@ -1583,7 +1579,6 @@ void call_rcu_sched(struct rcu_head *hea
}
EXPORT_SYMBOL_GPL(call_rcu_sched);

-#ifndef CONFIG_PREEMPT_RT_FULL
/*
* Queue an RCU for invocation after a quicker grace period.
*/
@@ -1592,7 +1587,6 @@ void call_rcu_bh(struct rcu_head *head,
__call_rcu(head, func, &rcu_bh_state);
}
EXPORT_SYMBOL_GPL(call_rcu_bh);
-#endif

/**
* synchronize_sched - wait until an rcu-sched grace period has elapsed.
@@ -1634,7 +1628,6 @@ void synchronize_sched(void)
}
EXPORT_SYMBOL_GPL(synchronize_sched);

-#ifndef CONFIG_PREEMPT_RT_FULL
/**
* synchronize_rcu_bh - wait until an rcu_bh grace period has elapsed.
*
@@ -1660,7 +1653,6 @@ void synchronize_rcu_bh(void)
destroy_rcu_head_on_stack(&rcu.head);
}
EXPORT_SYMBOL_GPL(synchronize_rcu_bh);
-#endif

/*
* Check to see if there is any immediate RCU-related work to be done
@@ -1814,7 +1806,6 @@ static void _rcu_barrier(struct rcu_stat
mutex_unlock(&rcu_barrier_mutex);
}

-#ifndef CONFIG_PREEMPT_RT_FULL
/**
* rcu_barrier_bh - Wait until all in-flight call_rcu_bh() callbacks complete.
*/
@@ -1823,7 +1814,6 @@ void rcu_barrier_bh(void)
_rcu_barrier(&rcu_bh_state, call_rcu_bh);
}
EXPORT_SYMBOL_GPL(rcu_barrier_bh);
-#endif

/**
* rcu_barrier_sched - Wait for in-flight call_rcu_sched() callbacks.
Index: linux-2.6/kernel/rtmutex-debug.c
===================================================================
--- linux-2.6.orig/kernel/rtmutex-debug.c
+++ linux-2.6/kernel/rtmutex-debug.c
@@ -94,8 +94,10 @@ void debug_rt_mutex_print_deadlock(struc
return;
}

- if (!debug_locks_off())
+ if (!debug_locks_off()) {
+ rcu_read_unlock();
return;
+ }

printk("\n============================================\n");
printk( "[ BUG: circular locking deadlock detected! ]\n");
Index: linux-2.6/kernel/rtmutex.c
===================================================================
--- linux-2.6.orig/kernel/rtmutex.c
+++ linux-2.6/kernel/rtmutex.c
@@ -659,7 +659,7 @@ static inline void rt_spin_lock_fastunlo
slowfn(lock);
}

-#ifdef CONFIG_SMP_X
+#ifdef CONFIG_SMP
/*
* Note that owner is a speculative pointer and dereferencing relies
* on rcu_read_lock() and the check against the lock owner.
Index: linux-2.6/kernel/sched.c
===================================================================
--- linux-2.6.orig/kernel/sched.c
+++ linux-2.6/kernel/sched.c
@@ -4207,6 +4207,126 @@ static inline void schedule_debug(struct
schedstat_inc(this_rq(), sched_count);
}

+#ifdef CONFIG_PREEMPT_RT_FULL
+#define MIGRATE_DISABLE_SET_AFFIN (1<<30) /* Can't make a negative */
+#define migrate_disabled_updated(p) ((p)->migrate_disable & MIGRATE_DISABLE_SET_AFFIN)
+#define migrate_disable_count(p) ((p)->migrate_disable & ~MIGRATE_DISABLE_SET_AFFIN)
+
+static inline void update_migrate_disable(struct task_struct *p)
+{
+ const struct cpumask *mask;
+
+ if (likely(!p->migrate_disable))
+ return;
+
+ /* Did we already update affinity? */
+ if (unlikely(migrate_disabled_updated(p)))
+ return;
+
+ /*
+ * Since this is always current we can get away with only locking
+ * rq->lock, the ->cpus_allowed value can normally only be changed
+ * while holding both p->pi_lock and rq->lock, but seeing that this
+ * is current, we cannot actually be waking up, so all code that
+ * relies on serialization against p->pi_lock is out of scope.
+ *
+ * Having rq->lock serializes us against things like
+ * set_cpus_allowed_ptr() that can still happen concurrently.
+ */
+ mask = tsk_cpus_allowed(p);
+
+ if (p->sched_class->set_cpus_allowed)
+ p->sched_class->set_cpus_allowed(p, mask);
+ p->rt.nr_cpus_allowed = cpumask_weight(mask);
+
+ /* Let migrate_enable know to fix things back up */
+ p->migrate_disable |= MIGRATE_DISABLE_SET_AFFIN;
+}
+
+void migrate_disable(void)
+{
+ struct task_struct *p = current;
+
+ if (in_atomic() || p->flags & PF_THREAD_BOUND) {
+#ifdef CONFIG_SCHED_DEBUG
+ p->migrate_disable_atomic++;
+#endif
+ return;
+ }
+
+#ifdef CONFIG_SCHED_DEBUG
+ WARN_ON_ONCE(p->migrate_disable_atomic);
+#endif
+
+ preempt_disable();
+ if (p->migrate_disable) {
+ p->migrate_disable++;
+ preempt_enable();
+ return;
+ }
+
+ pin_current_cpu();
+ p->migrate_disable = 1;
+ preempt_enable();
+}
+EXPORT_SYMBOL_GPL(migrate_disable);
+
+void migrate_enable(void)
+{
+ struct task_struct *p = current;
+ const struct cpumask *mask;
+ unsigned long flags;
+ struct rq *rq;
+
+ if (in_atomic() || p->flags & PF_THREAD_BOUND) {
+#ifdef CONFIG_SCHED_DEBUG
+ p->migrate_disable_atomic--;
+#endif
+ return;
+ }
+
+#ifdef CONFIG_SCHED_DEBUG
+ WARN_ON_ONCE(p->migrate_disable_atomic);
+#endif
+ WARN_ON_ONCE(p->migrate_disable <= 0);
+
+ preempt_disable();
+ if (migrate_disable_count(p) > 1) {
+ p->migrate_disable--;
+ preempt_enable();
+ return;
+ }
+
+ if (unlikely(migrate_disabled_updated(p))) {
+ /*
+ * Undo whatever update_migrate_disable() did, also see there
+ * about locking.
+ */
+ rq = this_rq();
+ raw_spin_lock_irqsave(&rq->lock, flags);
+
+ /*
+ * Clearing migrate_disable causes tsk_cpus_allowed to
+ * show the tasks original cpu affinity.
+ */
+ p->migrate_disable = 0;
+ mask = tsk_cpus_allowed(p);
+ if (p->sched_class->set_cpus_allowed)
+ p->sched_class->set_cpus_allowed(p, mask);
+ p->rt.nr_cpus_allowed = cpumask_weight(mask);
+ raw_spin_unlock_irqrestore(&rq->lock, flags);
+ } else
+ p->migrate_disable = 0;
+
+ unpin_current_cpu();
+ preempt_enable();
+}
+EXPORT_SYMBOL_GPL(migrate_enable);
+#else
+static inline void update_migrate_disable(struct task_struct *p) { }
+#define migrate_disabled_updated(p) 0
+#endif
+
static void put_prev_task(struct rq *rq, struct task_struct *prev)
{
if (prev->on_rq || rq->skip_clock_update < 0)
@@ -4266,6 +4386,8 @@ need_resched:

raw_spin_lock_irq(&rq->lock);

+ update_migrate_disable(prev);
+
switch_count = &prev->nivcsw;
if (prev->state && !(preempt_count() & PREEMPT_ACTIVE)) {
if (unlikely(signal_pending_state(prev->state, prev))) {
@@ -4433,7 +4555,16 @@ asmlinkage void __sched notrace preempt_

do {
add_preempt_count_notrace(PREEMPT_ACTIVE);
+ /*
+ * The add/subtract must not be traced by the function
+ * tracer. But we still want to account for the
+ * preempt off latency tracer. Since the _notrace versions
+ * of add/subtract skip the accounting for latency tracer
+ * we must force it manually.
+ */
+ start_critical_timings();
__schedule();
+ stop_critical_timings();
sub_preempt_count_notrace(PREEMPT_ACTIVE);

/*
@@ -6058,7 +6189,7 @@ static inline void sched_init_granularit
#ifdef CONFIG_SMP
void do_set_cpus_allowed(struct task_struct *p, const struct cpumask *new_mask)
{
- if (!__migrate_disabled(p)) {
+ if (!migrate_disabled_updated(p)) {
if (p->sched_class && p->sched_class->set_cpus_allowed)
p->sched_class->set_cpus_allowed(p, new_mask);
p->rt.nr_cpus_allowed = cpumask_weight(new_mask);
@@ -6133,124 +6264,6 @@ out:
}
EXPORT_SYMBOL_GPL(set_cpus_allowed_ptr);

-#ifdef CONFIG_PREEMPT_RT_FULL
-void migrate_disable(void)
-{
- struct task_struct *p = current;
- const struct cpumask *mask;
- unsigned long flags;
- struct rq *rq;
-
- if (in_atomic()) {
-#ifdef CONFIG_SCHED_DEBUG
- p->migrate_disable_atomic++;
-#endif
- return;
- }
-
-#ifdef CONFIG_SCHED_DEBUG
- WARN_ON_ONCE(p->migrate_disable_atomic);
-#endif
-
- preempt_disable();
- if (p->migrate_disable) {
- p->migrate_disable++;
- preempt_enable();
- return;
- }
-
- pin_current_cpu();
- if (unlikely(!scheduler_running)) {
- p->migrate_disable = 1;
- preempt_enable();
- return;
- }
-
- /*
- * Since this is always current we can get away with only locking
- * rq->lock, the ->cpus_allowed value can normally only be changed
- * while holding both p->pi_lock and rq->lock, but seeing that this
- * it current, we cannot actually be waking up, so all code that
- * relies on serialization against p->pi_lock is out of scope.
- *
- * Taking rq->lock serializes us against things like
- * set_cpus_allowed_ptr() that can still happen concurrently.
- */
- rq = this_rq();
- raw_spin_lock_irqsave(&rq->lock, flags);
- p->migrate_disable = 1;
- mask = tsk_cpus_allowed(p);
-
- WARN_ON(!cpumask_test_cpu(smp_processor_id(), mask));
-
- if (!cpumask_equal(&p->cpus_allowed, mask)) {
- if (p->sched_class->set_cpus_allowed)
- p->sched_class->set_cpus_allowed(p, mask);
- p->rt.nr_cpus_allowed = cpumask_weight(mask);
- }
- raw_spin_unlock_irqrestore(&rq->lock, flags);
- preempt_enable();
-}
-EXPORT_SYMBOL_GPL(migrate_disable);
-
-void migrate_enable(void)
-{
- struct task_struct *p = current;
- const struct cpumask *mask;
- unsigned long flags;
- struct rq *rq;
-
- if (in_atomic()) {
-#ifdef CONFIG_SCHED_DEBUG
- p->migrate_disable_atomic--;
-#endif
- return;
- }
-
-#ifdef CONFIG_SCHED_DEBUG
- WARN_ON_ONCE(p->migrate_disable_atomic);
-#endif
- WARN_ON_ONCE(p->migrate_disable <= 0);
-
- preempt_disable();
- if (p->migrate_disable > 1) {
- p->migrate_disable--;
- preempt_enable();
- return;
- }
-
- if (unlikely(!scheduler_running)) {
- p->migrate_disable = 0;
- unpin_current_cpu();
- preempt_enable();
- return;
- }
-
- /*
- * See comment in migrate_disable().
- */
- rq = this_rq();
- raw_spin_lock_irqsave(&rq->lock, flags);
- mask = tsk_cpus_allowed(p);
- p->migrate_disable = 0;
-
- WARN_ON(!cpumask_test_cpu(smp_processor_id(), mask));
-
- if (!cpumask_equal(&p->cpus_allowed, mask)) {
- /* Get the mask now that migration is enabled */
- mask = tsk_cpus_allowed(p);
- if (p->sched_class->set_cpus_allowed)
- p->sched_class->set_cpus_allowed(p, mask);
- p->rt.nr_cpus_allowed = cpumask_weight(mask);
- }
-
- raw_spin_unlock_irqrestore(&rq->lock, flags);
- unpin_current_cpu();
- preempt_enable();
-}
-EXPORT_SYMBOL_GPL(migrate_enable);
-#endif /* CONFIG_PREEMPT_RT_FULL */
-
/*
* Move (not current) task off this cpu, onto dest cpu. We're doing
* this because either it can't run here any more (set_cpus_allowed()
Index: linux-2.6/kernel/signal.c
===================================================================
--- linux-2.6.orig/kernel/signal.c
+++ linux-2.6/kernel/signal.c
@@ -1860,15 +1860,7 @@ static void ptrace_stop(int exit_code, i
if (gstop_done && !real_parent_is_ptracer(current))
do_notify_parent_cldstop(current, false, why);

- /*
- * Don't want to allow preemption here, because
- * sys_ptrace() needs this task to be inactive.
- *
- * XXX: implement read_unlock_no_resched().
- */
- preempt_disable();
read_unlock(&tasklist_lock);
- __preempt_enable_no_resched();
schedule();
} else {
/*
Index: linux-2.6/kernel/softirq.c
===================================================================
--- linux-2.6.orig/kernel/softirq.c
+++ linux-2.6/kernel/softirq.c
@@ -1104,9 +1104,8 @@ static int __cpuinit cpu_callback(struct
int hotcpu = (unsigned long)hcpu;
struct task_struct *p;

- switch (action) {
+ switch (action & ~CPU_TASKS_FROZEN) {
case CPU_UP_PREPARE:
- case CPU_UP_PREPARE_FROZEN:
p = kthread_create_on_node(run_ksoftirqd,
hcpu,
cpu_to_node(hotcpu),
@@ -1119,19 +1118,16 @@ static int __cpuinit cpu_callback(struct
per_cpu(ksoftirqd, hotcpu) = p;
break;
case CPU_ONLINE:
- case CPU_ONLINE_FROZEN:
wake_up_process(per_cpu(ksoftirqd, hotcpu));
break;
#ifdef CONFIG_HOTPLUG_CPU
case CPU_UP_CANCELED:
- case CPU_UP_CANCELED_FROZEN:
if (!per_cpu(ksoftirqd, hotcpu))
break;
/* Unbind so it can run. Fall thru. */
kthread_bind(per_cpu(ksoftirqd, hotcpu),
cpumask_any(cpu_online_mask));
- case CPU_DEAD:
- case CPU_DEAD_FROZEN: {
+ case CPU_POST_DEAD: {
static const struct sched_param param = {
.sched_priority = MAX_RT_PRIO-1
};
Index: linux-2.6/kernel/time/Kconfig
===================================================================
--- linux-2.6.orig/kernel/time/Kconfig
+++ linux-2.6/kernel/time/Kconfig
@@ -7,7 +7,6 @@ config TICK_ONESHOT
config NO_HZ
bool "Tickless System (Dynamic Ticks)"
depends on !ARCH_USES_GETTIMEOFFSET && GENERIC_CLOCKEVENTS
- depends on !PREEMPT_RT_FULL
select TICK_ONESHOT
help
This option enables a tickless system: timer interrupts will
Index: linux-2.6/kernel/trace/ring_buffer.c
===================================================================
--- linux-2.6.orig/kernel/trace/ring_buffer.c
+++ linux-2.6/kernel/trace/ring_buffer.c
@@ -478,7 +478,7 @@ struct ring_buffer_per_cpu {
int cpu;
atomic_t record_disabled;
struct ring_buffer *buffer;
- raw_spinlock_t reader_lock; /* serialize readers */
+ spinlock_t reader_lock; /* serialize readers */
arch_spinlock_t lock;
struct lock_class_key lock_key;
struct list_head *pages;
@@ -1040,6 +1040,44 @@ static int rb_allocate_pages(struct ring
return -ENOMEM;
}

+static inline int ok_to_lock(void)
+{
+ if (in_nmi())
+ return 0;
+#ifdef CONFIG_PREEMPT_RT_FULL
+ if (in_atomic())
+ return 0;
+#endif
+ return 1;
+}
+
+static int
+read_buffer_lock(struct ring_buffer_per_cpu *cpu_buffer,
+ unsigned long *flags)
+{
+ /*
+ * If an NMI die dumps out the content of the ring buffer
+ * do not grab locks. We also permanently disable the ring
+ * buffer too. A one time deal is all you get from reading
+ * the ring buffer from an NMI.
+ */
+ if (!ok_to_lock()) {
+ if (spin_trylock_irqsave(&cpu_buffer->reader_lock, *flags))
+ return 1;
+ tracing_off_permanent();
+ return 0;
+ }
+ spin_lock_irqsave(&cpu_buffer->reader_lock, *flags);
+ return 1;
+}
+
+static void
+read_buffer_unlock(struct ring_buffer_per_cpu *cpu_buffer,
+ unsigned long flags, int locked)
+{
+ if (locked)
+ spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
+}
static struct ring_buffer_per_cpu *
rb_allocate_cpu_buffer(struct ring_buffer *buffer, int cpu)
{
@@ -1055,7 +1093,7 @@ rb_allocate_cpu_buffer(struct ring_buffe

cpu_buffer->cpu = cpu;
cpu_buffer->buffer = buffer;
- raw_spin_lock_init(&cpu_buffer->reader_lock);
+ spin_lock_init(&cpu_buffer->reader_lock);
lockdep_set_class(&cpu_buffer->reader_lock, buffer->reader_lock_key);
cpu_buffer->lock = (arch_spinlock_t)__ARCH_SPIN_LOCK_UNLOCKED;

@@ -1250,9 +1288,11 @@ rb_remove_pages(struct ring_buffer_per_c
{
struct buffer_page *bpage;
struct list_head *p;
+ unsigned long flags;
unsigned i;
+ int locked;

- raw_spin_lock_irq(&cpu_buffer->reader_lock);
+ locked = read_buffer_lock(cpu_buffer, &flags);
rb_head_page_deactivate(cpu_buffer);

for (i = 0; i < nr_pages; i++) {
@@ -1270,7 +1310,7 @@ rb_remove_pages(struct ring_buffer_per_c
rb_check_pages(cpu_buffer);

out:
- raw_spin_unlock_irq(&cpu_buffer->reader_lock);
+ read_buffer_unlock(cpu_buffer, flags, locked);
}

static void
@@ -1279,9 +1319,11 @@ rb_insert_pages(struct ring_buffer_per_c
{
struct buffer_page *bpage;
struct list_head *p;
+ unsigned long flags;
unsigned i;
+ int locked;

- raw_spin_lock_irq(&cpu_buffer->reader_lock);
+ locked = read_buffer_lock(cpu_buffer, &flags);
rb_head_page_deactivate(cpu_buffer);

for (i = 0; i < nr_pages; i++) {
@@ -1296,7 +1338,7 @@ rb_insert_pages(struct ring_buffer_per_c
rb_check_pages(cpu_buffer);

out:
- raw_spin_unlock_irq(&cpu_buffer->reader_lock);
+ read_buffer_unlock(cpu_buffer, flags, locked);
}

/**
@@ -2784,15 +2826,16 @@ void ring_buffer_iter_reset(struct ring_
{
struct ring_buffer_per_cpu *cpu_buffer;
unsigned long flags;
+ int locked;

if (!iter)
return;

cpu_buffer = iter->cpu_buffer;

- raw_spin_lock_irqsave(&cpu_buffer->reader_lock, flags);
+ locked = read_buffer_lock(cpu_buffer, &flags);
rb_iter_reset(iter);
- raw_spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
+ read_buffer_unlock(cpu_buffer, flags, locked);
}
EXPORT_SYMBOL_GPL(ring_buffer_iter_reset);

@@ -3210,21 +3253,6 @@ rb_iter_peek(struct ring_buffer_iter *it
}
EXPORT_SYMBOL_GPL(ring_buffer_iter_peek);

-static inline int rb_ok_to_lock(void)
-{
- /*
- * If an NMI die dumps out the content of the ring buffer
- * do not grab locks. We also permanently disable the ring
- * buffer too. A one time deal is all you get from reading
- * the ring buffer from an NMI.
- */
- if (likely(!in_nmi()))
- return 1;
-
- tracing_off_permanent();
- return 0;
-}
-
/**
* ring_buffer_peek - peek at the next event to be read
* @buffer: The ring buffer to read
@@ -3242,22 +3270,17 @@ ring_buffer_peek(struct ring_buffer *buf
struct ring_buffer_per_cpu *cpu_buffer = buffer->buffers[cpu];
struct ring_buffer_event *event;
unsigned long flags;
- int dolock;
+ int locked;

if (!cpumask_test_cpu(cpu, buffer->cpumask))
return NULL;

- dolock = rb_ok_to_lock();
again:
- local_irq_save(flags);
- if (dolock)
- raw_spin_lock(&cpu_buffer->reader_lock);
+ locked = read_buffer_lock(cpu_buffer, &flags);
event = rb_buffer_peek(cpu_buffer, ts, lost_events);
if (event && event->type_len == RINGBUF_TYPE_PADDING)
rb_advance_reader(cpu_buffer);
- if (dolock)
- raw_spin_unlock(&cpu_buffer->reader_lock);
- local_irq_restore(flags);
+ read_buffer_unlock(cpu_buffer, flags, locked);

if (event && event->type_len == RINGBUF_TYPE_PADDING)
goto again;
@@ -3279,11 +3302,12 @@ ring_buffer_iter_peek(struct ring_buffer
struct ring_buffer_per_cpu *cpu_buffer = iter->cpu_buffer;
struct ring_buffer_event *event;
unsigned long flags;
+ int locked;

again:
- raw_spin_lock_irqsave(&cpu_buffer->reader_lock, flags);
+ locked = read_buffer_lock(cpu_buffer, &flags);
event = rb_iter_peek(iter, ts);
- raw_spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
+ read_buffer_unlock(cpu_buffer, flags, locked);

if (event && event->type_len == RINGBUF_TYPE_PADDING)
goto again;
@@ -3309,9 +3333,7 @@ ring_buffer_consume(struct ring_buffer *
struct ring_buffer_per_cpu *cpu_buffer;
struct ring_buffer_event *event = NULL;
unsigned long flags;
- int dolock;
-
- dolock = rb_ok_to_lock();
+ int locked;

again:
/* might be called in atomic */
@@ -3321,9 +3343,7 @@ ring_buffer_consume(struct ring_buffer *
goto out;

cpu_buffer = buffer->buffers[cpu];
- local_irq_save(flags);
- if (dolock)
- raw_spin_lock(&cpu_buffer->reader_lock);
+ locked = read_buffer_lock(cpu_buffer, &flags);

event = rb_buffer_peek(cpu_buffer, ts, lost_events);
if (event) {
@@ -3331,9 +3351,8 @@ ring_buffer_consume(struct ring_buffer *
rb_advance_reader(cpu_buffer);
}

- if (dolock)
- raw_spin_unlock(&cpu_buffer->reader_lock);
- local_irq_restore(flags);
+ read_buffer_unlock(cpu_buffer, flags, locked);
+

out:
preempt_enable();
@@ -3418,17 +3437,18 @@ ring_buffer_read_start(struct ring_buffe
{
struct ring_buffer_per_cpu *cpu_buffer;
unsigned long flags;
+ int locked;

if (!iter)
return;

cpu_buffer = iter->cpu_buffer;

- raw_spin_lock_irqsave(&cpu_buffer->reader_lock, flags);
+ locked = read_buffer_lock(cpu_buffer, &flags);
arch_spin_lock(&cpu_buffer->lock);
rb_iter_reset(iter);
arch_spin_unlock(&cpu_buffer->lock);
- raw_spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
+ read_buffer_unlock(cpu_buffer, flags, locked);
}
EXPORT_SYMBOL_GPL(ring_buffer_read_start);

@@ -3462,8 +3482,9 @@ ring_buffer_read(struct ring_buffer_iter
struct ring_buffer_event *event;
struct ring_buffer_per_cpu *cpu_buffer = iter->cpu_buffer;
unsigned long flags;
+ int locked;

- raw_spin_lock_irqsave(&cpu_buffer->reader_lock, flags);
+ locked = read_buffer_lock(cpu_buffer, &flags);
again:
event = rb_iter_peek(iter, ts);
if (!event)
@@ -3474,7 +3495,7 @@ ring_buffer_read(struct ring_buffer_iter

rb_advance_iter(iter);
out:
- raw_spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
+ read_buffer_unlock(cpu_buffer, flags, locked);

return event;
}
@@ -3537,13 +3558,14 @@ void ring_buffer_reset_cpu(struct ring_b
{
struct ring_buffer_per_cpu *cpu_buffer = buffer->buffers[cpu];
unsigned long flags;
+ int locked;

if (!cpumask_test_cpu(cpu, buffer->cpumask))
return;

atomic_inc(&cpu_buffer->record_disabled);

- raw_spin_lock_irqsave(&cpu_buffer->reader_lock, flags);
+ locked = read_buffer_lock(cpu_buffer, &flags);

if (RB_WARN_ON(cpu_buffer, local_read(&cpu_buffer->committing)))
goto out;
@@ -3555,7 +3577,7 @@ void ring_buffer_reset_cpu(struct ring_b
arch_spin_unlock(&cpu_buffer->lock);

out:
- raw_spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
+ read_buffer_unlock(cpu_buffer, flags, locked);

atomic_dec(&cpu_buffer->record_disabled);
}
@@ -3582,22 +3604,16 @@ int ring_buffer_empty(struct ring_buffer
{
struct ring_buffer_per_cpu *cpu_buffer;
unsigned long flags;
- int dolock;
+ int locked;
int cpu;
int ret;

- dolock = rb_ok_to_lock();
-
/* yes this is racy, but if you don't like the race, lock the buffer */
for_each_buffer_cpu(buffer, cpu) {
cpu_buffer = buffer->buffers[cpu];
- local_irq_save(flags);
- if (dolock)
- raw_spin_lock(&cpu_buffer->reader_lock);
+ locked = read_buffer_lock(cpu_buffer, &flags);
ret = rb_per_cpu_empty(cpu_buffer);
- if (dolock)
- raw_spin_unlock(&cpu_buffer->reader_lock);
- local_irq_restore(flags);
+ read_buffer_unlock(cpu_buffer, flags, locked);

if (!ret)
return 0;
@@ -3616,22 +3632,16 @@ int ring_buffer_empty_cpu(struct ring_bu
{
struct ring_buffer_per_cpu *cpu_buffer;
unsigned long flags;
- int dolock;
+ int locked;
int ret;

if (!cpumask_test_cpu(cpu, buffer->cpumask))
return 1;

- dolock = rb_ok_to_lock();
-
cpu_buffer = buffer->buffers[cpu];
- local_irq_save(flags);
- if (dolock)
- raw_spin_lock(&cpu_buffer->reader_lock);
+ locked = read_buffer_lock(cpu_buffer, &flags);
ret = rb_per_cpu_empty(cpu_buffer);
- if (dolock)
- raw_spin_unlock(&cpu_buffer->reader_lock);
- local_irq_restore(flags);
+ read_buffer_unlock(cpu_buffer, flags, locked);

return ret;
}
@@ -3805,6 +3815,7 @@ int ring_buffer_read_page(struct ring_bu
unsigned int commit;
unsigned int read;
u64 save_timestamp;
+ int locked;
int ret = -1;

if (!cpumask_test_cpu(cpu, buffer->cpumask))
@@ -3826,7 +3837,7 @@ int ring_buffer_read_page(struct ring_bu
if (!bpage)
goto out;

- raw_spin_lock_irqsave(&cpu_buffer->reader_lock, flags);
+ locked = read_buffer_lock(cpu_buffer, &flags);

reader = rb_get_reader_page(cpu_buffer);
if (!reader)
@@ -3949,7 +3960,7 @@ int ring_buffer_read_page(struct ring_bu
memset(&bpage->data[commit], 0, BUF_PAGE_SIZE - commit);

out_unlock:
- raw_spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
+ read_buffer_unlock(cpu_buffer, flags, locked);

out:
return ret;
Index: linux-2.6/kernel/trace/trace_irqsoff.c
===================================================================
--- linux-2.6.orig/kernel/trace/trace_irqsoff.c
+++ linux-2.6/kernel/trace/trace_irqsoff.c
@@ -513,14 +513,14 @@ EXPORT_SYMBOL(trace_hardirqs_off_caller)
void trace_preempt_on(unsigned long a0, unsigned long a1)
{
trace_preemptirqsoff_hist(PREEMPT_ON, 0);
- if (preempt_trace())
+ if (preempt_trace() && !irq_trace())
stop_critical_timing(a0, a1);
}

void trace_preempt_off(unsigned long a0, unsigned long a1)
{
- trace_preemptirqsoff_hist(PREEMPT_OFF, 1);
- if (preempt_trace())
+ trace_preemptirqsoff_hist(PREEMPT_ON, 1);
+ if (preempt_trace() && !irq_trace())
start_critical_timing(a0, a1);
}
#endif /* CONFIG_PREEMPT_TRACER */
Index: linux-2.6/kernel/workqueue.c
===================================================================
--- linux-2.6.orig/kernel/workqueue.c
+++ linux-2.6/kernel/workqueue.c
@@ -41,6 +41,7 @@
#include <linux/debug_locks.h>
#include <linux/lockdep.h>
#include <linux/idr.h>
+#include <linux/delay.h>

#include "workqueue_sched.h"

@@ -57,20 +58,10 @@ enum {
WORKER_DIE = 1 << 1, /* die die die */
WORKER_IDLE = 1 << 2, /* is idle */
WORKER_PREP = 1 << 3, /* preparing to run works */
- WORKER_ROGUE = 1 << 4, /* not bound to any cpu */
- WORKER_REBIND = 1 << 5, /* mom is home, come back */
- WORKER_CPU_INTENSIVE = 1 << 6, /* cpu intensive */
- WORKER_UNBOUND = 1 << 7, /* worker is unbound */
-
- WORKER_NOT_RUNNING = WORKER_PREP | WORKER_ROGUE | WORKER_REBIND |
- WORKER_CPU_INTENSIVE | WORKER_UNBOUND,
-
- /* gcwq->trustee_state */
- TRUSTEE_START = 0, /* start */
- TRUSTEE_IN_CHARGE = 1, /* trustee in charge of gcwq */
- TRUSTEE_BUTCHER = 2, /* butcher workers */
- TRUSTEE_RELEASE = 3, /* release workers */
- TRUSTEE_DONE = 4, /* trustee is done */
+ WORKER_CPU_INTENSIVE = 1 << 4, /* cpu intensive */
+ WORKER_UNBOUND = 1 << 5, /* worker is unbound */
+
+ WORKER_NOT_RUNNING = WORKER_PREP | WORKER_CPU_INTENSIVE | WORKER_UNBOUND,

BUSY_WORKER_HASH_ORDER = 6, /* 64 pointers */
BUSY_WORKER_HASH_SIZE = 1 << BUSY_WORKER_HASH_ORDER,
@@ -84,7 +75,6 @@ enum {
(min two ticks) */
MAYDAY_INTERVAL = HZ / 10, /* and then every 100ms */
CREATE_COOLDOWN = HZ, /* time to breath after fail */
- TRUSTEE_COOLDOWN = HZ / 10, /* for trustee draining */

/*
* Rescue workers are used only on emergencies and shared by
@@ -136,7 +126,6 @@ struct worker {
unsigned long last_active; /* L: last active timestamp */
unsigned int flags; /* X: flags */
int id; /* I: worker id */
- struct work_struct rebind_work; /* L: rebind worker to cpu */
int sleeping; /* None */
};

@@ -164,10 +153,8 @@ struct global_cwq {

struct ida worker_ida; /* L: for worker IDs */

- struct task_struct *trustee; /* L: for gcwq shutdown */
- unsigned int trustee_state; /* L: trustee state */
- wait_queue_head_t trustee_wait; /* trustee wait */
struct worker *first_idle; /* L: first idle worker */
+ wait_queue_head_t idle_wait;
} ____cacheline_aligned_in_smp;

/*
@@ -971,13 +958,38 @@ static bool is_chained_work(struct workq
return false;
}

-static void __queue_work(unsigned int cpu, struct workqueue_struct *wq,
- struct work_struct *work)
+static void ___queue_work(struct workqueue_struct *wq, struct global_cwq *gcwq,
+ struct work_struct *work)
{
- struct global_cwq *gcwq;
struct cpu_workqueue_struct *cwq;
struct list_head *worklist;
unsigned int work_flags;
+
+ /* gcwq determined, get cwq and queue */
+ cwq = get_cwq(gcwq->cpu, wq);
+ trace_workqueue_queue_work(gcwq->cpu, cwq, work);
+
+ BUG_ON(!list_empty(&work->entry));
+
+ cwq->nr_in_flight[cwq->work_color]++;
+ work_flags = work_color_to_flags(cwq->work_color);
+
+ if (likely(cwq->nr_active < cwq->max_active)) {
+ trace_workqueue_activate_work(work);
+ cwq->nr_active++;
+ worklist = gcwq_determine_ins_pos(gcwq, cwq);
+ } else {
+ work_flags |= WORK_STRUCT_DELAYED;
+ worklist = &cwq->delayed_works;
+ }
+
+ insert_work(cwq, work, worklist, work_flags);
+}
+
+static void __queue_work(unsigned int cpu, struct workqueue_struct *wq,
+ struct work_struct *work)
+{
+ struct global_cwq *gcwq;
unsigned long flags;

debug_work_activate(work);
@@ -1023,27 +1035,32 @@ static void __queue_work(unsigned int cp
spin_lock_irqsave(&gcwq->lock, flags);
}

- /* gcwq determined, get cwq and queue */
- cwq = get_cwq(gcwq->cpu, wq);
- trace_workqueue_queue_work(cpu, cwq, work);
+ ___queue_work(wq, gcwq, work);

- BUG_ON(!list_empty(&work->entry));
+ spin_unlock_irqrestore(&gcwq->lock, flags);
+}

- cwq->nr_in_flight[cwq->work_color]++;
- work_flags = work_color_to_flags(cwq->work_color);
+/**
+ * queue_work_on - queue work on specific cpu
+ * @cpu: CPU number to execute work on
+ * @wq: workqueue to use
+ * @work: work to queue
+ *
+ * Returns 0 if @work was already on a queue, non-zero otherwise.
+ *
+ * We queue the work to a specific CPU, the caller must ensure it
+ * can't go away.
+ */
+static int
+__queue_work_on(int cpu, struct workqueue_struct *wq, struct work_struct *work)
+{
+ int ret = 0;

- if (likely(cwq->nr_active < cwq->max_active)) {
- trace_workqueue_activate_work(work);
- cwq->nr_active++;
- worklist = gcwq_determine_ins_pos(gcwq, cwq);
- } else {
- work_flags |= WORK_STRUCT_DELAYED;
- worklist = &cwq->delayed_works;
+ if (!test_and_set_bit(WORK_STRUCT_PENDING_BIT, work_data_bits(work))) {
+ __queue_work(cpu, wq, work);
+ ret = 1;
}
-
- insert_work(cwq, work, worklist, work_flags);
-
- spin_unlock_irqrestore(&gcwq->lock, flags);
+ return ret;
}

/**
@@ -1060,34 +1077,19 @@ int queue_work(struct workqueue_struct *
{
int ret;

- ret = queue_work_on(get_cpu_light(), wq, work);
+ ret = __queue_work_on(get_cpu_light(), wq, work);
put_cpu_light();

return ret;
}
EXPORT_SYMBOL_GPL(queue_work);

-/**
- * queue_work_on - queue work on specific cpu
- * @cpu: CPU number to execute work on
- * @wq: workqueue to use
- * @work: work to queue
- *
- * Returns 0 if @work was already on a queue, non-zero otherwise.
- *
- * We queue the work to a specific CPU, the caller must ensure it
- * can't go away.
- */
int
queue_work_on(int cpu, struct workqueue_struct *wq, struct work_struct *work)
{
- int ret = 0;
+ WARN_ON(wq->flags & WQ_NON_AFFINE);

- if (!test_and_set_bit(WORK_STRUCT_PENDING_BIT, work_data_bits(work))) {
- __queue_work(cpu, wq, work);
- ret = 1;
- }
- return ret;
+ return __queue_work_on(cpu, wq, work);
}
EXPORT_SYMBOL_GPL(queue_work_on);

@@ -1133,6 +1135,8 @@ int queue_delayed_work_on(int cpu, struc
struct timer_list *timer = &dwork->timer;
struct work_struct *work = &dwork->work;

+ WARN_ON((wq->flags & WQ_NON_AFFINE) && cpu != -1);
+
if (!test_and_set_bit(WORK_STRUCT_PENDING_BIT, work_data_bits(work))) {
unsigned int lcpu;

@@ -1198,12 +1202,13 @@ static void worker_enter_idle(struct wor
/* idle_list is LIFO */
list_add(&worker->entry, &gcwq->idle_list);

- if (likely(!(worker->flags & WORKER_ROGUE))) {
- if (too_many_workers(gcwq) && !timer_pending(&gcwq->idle_timer))
- mod_timer(&gcwq->idle_timer,
- jiffies + IDLE_WORKER_TIMEOUT);
- } else
- wake_up_all(&gcwq->trustee_wait);
+ if (gcwq->nr_idle == gcwq->nr_workers)
+ wake_up_all(&gcwq->idle_wait);
+
+ if (too_many_workers(gcwq) && !timer_pending(&gcwq->idle_timer)) {
+ mod_timer(&gcwq->idle_timer,
+ jiffies + IDLE_WORKER_TIMEOUT);
+ }

/* sanity check nr_running */
WARN_ON_ONCE(gcwq->nr_workers == gcwq->nr_idle &&
@@ -1272,8 +1277,14 @@ __acquires(&gcwq->lock)
* it races with cpu hotunplug operation. Verify
* against GCWQ_DISASSOCIATED.
*/
- if (!(gcwq->flags & GCWQ_DISASSOCIATED))
+ if (!(gcwq->flags & GCWQ_DISASSOCIATED)) {
+ /*
+ * Since we're binding to a particular cpu and need to
+ * stay there for correctness, mark us PF_THREAD_BOUND.
+ */
+ task->flags |= PF_THREAD_BOUND;
set_cpus_allowed_ptr(task, get_cpu_mask(gcwq->cpu));
+ }

spin_lock_irq(&gcwq->lock);
if (gcwq->flags & GCWQ_DISASSOCIATED)
@@ -1295,20 +1306,15 @@ __acquires(&gcwq->lock)
}
}

-/*
- * Function for worker->rebind_work used to rebind rogue busy workers
- * to the associated cpu which is coming back online. This is
- * scheduled by cpu up but can race with other cpu hotplug operations
- * and may be executed twice without intervening cpu down.
- */
-static void worker_rebind_fn(struct work_struct *work)
+static void worker_unbind_and_unlock(struct worker *worker)
{
- struct worker *worker = container_of(work, struct worker, rebind_work);
struct global_cwq *gcwq = worker->gcwq;
+ struct task_struct *task = worker->task;

- if (worker_maybe_bind_and_lock(worker))
- worker_clr_flags(worker, WORKER_REBIND);
-
+ /*
+ * Its no longer required we're PF_THREAD_BOUND, the work is done.
+ */
+ task->flags &= ~PF_THREAD_BOUND;
spin_unlock_irq(&gcwq->lock);
}

@@ -1320,7 +1326,6 @@ static struct worker *alloc_worker(void)
if (worker) {
INIT_LIST_HEAD(&worker->entry);
INIT_LIST_HEAD(&worker->scheduled);
- INIT_WORK(&worker->rebind_work, worker_rebind_fn);
/* on creation a worker is in !idle && prep state */
worker->flags = WORKER_PREP;
}
@@ -1375,15 +1380,9 @@ static struct worker *create_worker(stru
if (IS_ERR(worker->task))
goto fail;

- /*
- * A rogue worker will become a regular one if CPU comes
- * online later on. Make sure every worker has
- * PF_THREAD_BOUND set.
- */
if (bind && !on_unbound_cpu)
kthread_bind(worker->task, gcwq->cpu);
else {
- worker->task->flags |= PF_THREAD_BOUND;
if (on_unbound_cpu)
worker->flags |= WORKER_UNBOUND;
}
@@ -1660,13 +1659,6 @@ static bool manage_workers(struct worker

gcwq->flags &= ~GCWQ_MANAGING_WORKERS;

- /*
- * The trustee might be waiting to take over the manager
- * position, tell it we're done.
- */
- if (unlikely(gcwq->trustee))
- wake_up_all(&gcwq->trustee_wait);
-
return ret;
}

@@ -2067,7 +2059,7 @@ repeat:
if (keep_working(gcwq))
wake_up_worker(gcwq);

- spin_unlock_irq(&gcwq->lock);
+ worker_unbind_and_unlock(rescuer);
}

schedule();
@@ -2963,7 +2955,6 @@ struct workqueue_struct *__alloc_workque
if (IS_ERR(rescuer->task))
goto err;

- rescuer->task->flags |= PF_THREAD_BOUND;
wake_up_process(rescuer->task);
}

@@ -3177,171 +3168,71 @@ EXPORT_SYMBOL_GPL(work_busy);
* gcwqs serve mix of short, long and very long running works making
* blocked draining impractical.
*
- * This is solved by allowing a gcwq to be detached from CPU, running
- * it with unbound (rogue) workers and allowing it to be reattached
- * later if the cpu comes back online. A separate thread is created
- * to govern a gcwq in such state and is called the trustee of the
- * gcwq.
- *
- * Trustee states and their descriptions.
- *
- * START Command state used on startup. On CPU_DOWN_PREPARE, a
- * new trustee is started with this state.
- *
- * IN_CHARGE Once started, trustee will enter this state after
- * assuming the manager role and making all existing
- * workers rogue. DOWN_PREPARE waits for trustee to
- * enter this state. After reaching IN_CHARGE, trustee
- * tries to execute the pending worklist until it's empty
- * and the state is set to BUTCHER, or the state is set
- * to RELEASE.
- *
- * BUTCHER Command state which is set by the cpu callback after
- * the cpu has went down. Once this state is set trustee
- * knows that there will be no new works on the worklist
- * and once the worklist is empty it can proceed to
- * killing idle workers.
- *
- * RELEASE Command state which is set by the cpu callback if the
- * cpu down has been canceled or it has come online
- * again. After recognizing this state, trustee stops
- * trying to drain or butcher and clears ROGUE, rebinds
- * all remaining workers back to the cpu and releases
- * manager role.
- *
- * DONE Trustee will enter this state after BUTCHER or RELEASE
- * is complete.
- *
- * trustee CPU draining
- * took over down complete
- * START -----------> IN_CHARGE -----------> BUTCHER -----------> DONE
- * | | ^
- * | CPU is back online v return workers |
- * ----------------> RELEASE --------------
*/

-/**
- * trustee_wait_event_timeout - timed event wait for trustee
- * @cond: condition to wait for
- * @timeout: timeout in jiffies
- *
- * wait_event_timeout() for trustee to use. Handles locking and
- * checks for RELEASE request.
- *
- * CONTEXT:
- * spin_lock_irq(gcwq->lock) which may be released and regrabbed
- * multiple times. To be used by trustee.
- *
- * RETURNS:
- * Positive indicating left time if @cond is satisfied, 0 if timed
- * out, -1 if canceled.
- */
-#define trustee_wait_event_timeout(cond, timeout) ({ \
- long __ret = (timeout); \
- while (!((cond) || (gcwq->trustee_state == TRUSTEE_RELEASE)) && \
- __ret) { \
- spin_unlock_irq(&gcwq->lock); \
- __wait_event_timeout(gcwq->trustee_wait, (cond) || \
- (gcwq->trustee_state == TRUSTEE_RELEASE), \
- __ret); \
- spin_lock_irq(&gcwq->lock); \
- } \
- gcwq->trustee_state == TRUSTEE_RELEASE ? -1 : (__ret); \
-})
+static int __devinit workqueue_cpu_up_callback(struct notifier_block *nfb,
+ unsigned long action,
+ void *hcpu)
+{
+ unsigned int cpu = (unsigned long)hcpu;
+ struct global_cwq *gcwq = get_gcwq(cpu);
+ struct worker *uninitialized_var(new_worker);
+ unsigned long flags;

-/**
- * trustee_wait_event - event wait for trustee
- * @cond: condition to wait for
- *
- * wait_event() for trustee to use. Automatically handles locking and
- * checks for CANCEL request.
- *
- * CONTEXT:
- * spin_lock_irq(gcwq->lock) which may be released and regrabbed
- * multiple times. To be used by trustee.
- *
- * RETURNS:
- * 0 if @cond is satisfied, -1 if canceled.
- */
-#define trustee_wait_event(cond) ({ \
- long __ret1; \
- __ret1 = trustee_wait_event_timeout(cond, MAX_SCHEDULE_TIMEOUT);\
- __ret1 < 0 ? -1 : 0; \
-})
+ action &= ~CPU_TASKS_FROZEN;

-static int __cpuinit trustee_thread(void *__gcwq)
-{
- struct global_cwq *gcwq = __gcwq;
- struct worker *worker;
- struct work_struct *work;
- struct hlist_node *pos;
- long rc;
- int i;
+ switch (action) {
+ case CPU_UP_PREPARE:
+ BUG_ON(gcwq->first_idle);
+ new_worker = create_worker(gcwq, false);
+ if (!new_worker)
+ return NOTIFY_BAD;
+ }

- BUG_ON(gcwq->cpu != smp_processor_id());
+ /* some are called w/ irq disabled, don't disturb irq status */
+ spin_lock_irqsave(&gcwq->lock, flags);

- spin_lock_irq(&gcwq->lock);
- /*
- * Claim the manager position and make all workers rogue.
- * Trustee must be bound to the target cpu and can't be
- * cancelled.
- */
- BUG_ON(gcwq->cpu != smp_processor_id());
- rc = trustee_wait_event(!(gcwq->flags & GCWQ_MANAGING_WORKERS));
- BUG_ON(rc < 0);
+ switch (action) {
+ case CPU_UP_PREPARE:
+ BUG_ON(gcwq->first_idle);
+ gcwq->first_idle = new_worker;
+ break;

- gcwq->flags |= GCWQ_MANAGING_WORKERS;
+ case CPU_UP_CANCELED:
+ destroy_worker(gcwq->first_idle);
+ gcwq->first_idle = NULL;
+ break;

- list_for_each_entry(worker, &gcwq->idle_list, entry)
- worker->flags |= WORKER_ROGUE;
+ case CPU_ONLINE:
+ spin_unlock_irq(&gcwq->lock);
+ kthread_bind(gcwq->first_idle->task, cpu);
+ spin_lock_irq(&gcwq->lock);
+ gcwq->flags |= GCWQ_MANAGE_WORKERS;
+ start_worker(gcwq->first_idle);
+ gcwq->first_idle = NULL;
+ break;
+ }

- for_each_busy_worker(worker, i, pos, gcwq)
- worker->flags |= WORKER_ROGUE;
+ spin_unlock_irqrestore(&gcwq->lock, flags);

- /*
- * Call schedule() so that we cross rq->lock and thus can
- * guarantee sched callbacks see the rogue flag. This is
- * necessary as scheduler callbacks may be invoked from other
- * cpus.
- */
- spin_unlock_irq(&gcwq->lock);
- schedule();
- spin_lock_irq(&gcwq->lock);
+ return notifier_from_errno(0);
+}

- /*
- * Sched callbacks are disabled now. Zap nr_running. After
- * this, nr_running stays zero and need_more_worker() and
- * keep_working() are always true as long as the worklist is
- * not empty.
- */
- atomic_set(get_gcwq_nr_running(gcwq->cpu), 0);
+static void flush_gcwq(struct global_cwq *gcwq)
+{
+ struct work_struct *work, *nw;
+ struct worker *worker, *n;
+ LIST_HEAD(non_affine_works);

- spin_unlock_irq(&gcwq->lock);
- del_timer_sync(&gcwq->idle_timer);
spin_lock_irq(&gcwq->lock);
+ list_for_each_entry_safe(work, nw, &gcwq->worklist, entry) {
+ struct workqueue_struct *wq = get_work_cwq(work)->wq;

- /*
- * We're now in charge. Notify and proceed to drain. We need
- * to keep the gcwq running during the whole CPU down
- * procedure as other cpu hotunplug callbacks may need to
- * flush currently running tasks.
- */
- gcwq->trustee_state = TRUSTEE_IN_CHARGE;
- wake_up_all(&gcwq->trustee_wait);
-
- /*
- * The original cpu is in the process of dying and may go away
- * anytime now. When that happens, we and all workers would
- * be migrated to other cpus. Try draining any left work. We
- * want to get it over with ASAP - spam rescuers, wake up as
- * many idlers as necessary and create new ones till the
- * worklist is empty. Note that if the gcwq is frozen, there
- * may be frozen works in freezable cwqs. Don't declare
- * completion while frozen.
- */
- while (gcwq->nr_workers != gcwq->nr_idle ||
- gcwq->flags & GCWQ_FREEZING ||
- gcwq->trustee_state == TRUSTEE_IN_CHARGE) {
+ if (wq->flags & WQ_NON_AFFINE)
+ list_move(&work->entry, &non_affine_works);
+ }
+
+ while (!list_empty(&gcwq->worklist)) {
int nr_works = 0;

list_for_each_entry(work, &gcwq->worklist, entry) {
@@ -3355,200 +3246,55 @@ static int __cpuinit trustee_thread(void
wake_up_process(worker->task);
}

+ spin_unlock_irq(&gcwq->lock);
+
if (need_to_create_worker(gcwq)) {
- spin_unlock_irq(&gcwq->lock);
- worker = create_worker(gcwq, false);
- spin_lock_irq(&gcwq->lock);
- if (worker) {
- worker->flags |= WORKER_ROGUE;
+ worker = create_worker(gcwq, true);
+ if (worker)
start_worker(worker);
- }
}

- /* give a breather */
- if (trustee_wait_event_timeout(false, TRUSTEE_COOLDOWN) < 0)
- break;
- }
-
- /*
- * Either all works have been scheduled and cpu is down, or
- * cpu down has already been canceled. Wait for and butcher
- * all workers till we're canceled.
- */
- do {
- rc = trustee_wait_event(!list_empty(&gcwq->idle_list));
- while (!list_empty(&gcwq->idle_list))
- destroy_worker(list_first_entry(&gcwq->idle_list,
- struct worker, entry));
- } while (gcwq->nr_workers && rc >= 0);
-
- /*
- * At this point, either draining has completed and no worker
- * is left, or cpu down has been canceled or the cpu is being
- * brought back up. There shouldn't be any idle one left.
- * Tell the remaining busy ones to rebind once it finishes the
- * currently scheduled works by scheduling the rebind_work.
- */
- WARN_ON(!list_empty(&gcwq->idle_list));
+ wait_event_timeout(gcwq->idle_wait,
+ gcwq->nr_idle == gcwq->nr_workers, HZ/10);

- for_each_busy_worker(worker, i, pos, gcwq) {
- struct work_struct *rebind_work = &worker->rebind_work;
+ spin_lock_irq(&gcwq->lock);
+ }

- /*
- * Rebind_work may race with future cpu hotplug
- * operations. Use a separate flag to mark that
- * rebinding is scheduled.
- */
- worker->flags |= WORKER_REBIND;
- worker->flags &= ~WORKER_ROGUE;
+ WARN_ON(gcwq->nr_workers != gcwq->nr_idle);

- /* queue rebind_work, wq doesn't matter, use the default one */
- if (test_and_set_bit(WORK_STRUCT_PENDING_BIT,
- work_data_bits(rebind_work)))
- continue;
+ list_for_each_entry_safe(worker, n, &gcwq->idle_list, entry)
+ destroy_worker(worker);

- debug_work_activate(rebind_work);
- insert_work(get_cwq(gcwq->cpu, system_wq), rebind_work,
- worker->scheduled.next,
- work_color_to_flags(WORK_NO_COLOR));
- }
+ WARN_ON(gcwq->nr_workers || gcwq->nr_idle);

- /* relinquish manager role */
- gcwq->flags &= ~GCWQ_MANAGING_WORKERS;
-
- /* notify completion */
- gcwq->trustee = NULL;
- gcwq->trustee_state = TRUSTEE_DONE;
- wake_up_all(&gcwq->trustee_wait);
spin_unlock_irq(&gcwq->lock);
- return 0;
-}

-/**
- * wait_trustee_state - wait for trustee to enter the specified state
- * @gcwq: gcwq the trustee of interest belongs to
- * @state: target state to wait for
- *
- * Wait for the trustee to reach @state. DONE is already matched.
- *
- * CONTEXT:
- * spin_lock_irq(gcwq->lock) which may be released and regrabbed
- * multiple times. To be used by cpu_callback.
- */
-static void __cpuinit wait_trustee_state(struct global_cwq *gcwq, int state)
-__releases(&gcwq->lock)
-__acquires(&gcwq->lock)
-{
- if (!(gcwq->trustee_state == state ||
- gcwq->trustee_state == TRUSTEE_DONE)) {
- spin_unlock_irq(&gcwq->lock);
- __wait_event(gcwq->trustee_wait,
- gcwq->trustee_state == state ||
- gcwq->trustee_state == TRUSTEE_DONE);
- spin_lock_irq(&gcwq->lock);
+ gcwq = get_gcwq(get_cpu());
+ spin_lock_irq(&gcwq->lock);
+ list_for_each_entry_safe(work, nw, &non_affine_works, entry) {
+ list_del_init(&work->entry);
+ ___queue_work(get_work_cwq(work)->wq, gcwq, work);
}
+ spin_unlock_irq(&gcwq->lock);
+ put_cpu();
}

-static int __devinit workqueue_cpu_callback(struct notifier_block *nfb,
+static int __devinit workqueue_cpu_down_callback(struct notifier_block *nfb,
unsigned long action,
void *hcpu)
{
unsigned int cpu = (unsigned long)hcpu;
struct global_cwq *gcwq = get_gcwq(cpu);
- struct task_struct *new_trustee = NULL;
- struct worker *uninitialized_var(new_worker);
- unsigned long flags;

action &= ~CPU_TASKS_FROZEN;

- switch (action) {
- case CPU_DOWN_PREPARE:
- new_trustee = kthread_create(trustee_thread, gcwq,
- "workqueue_trustee/%d\n", cpu);
- if (IS_ERR(new_trustee))
- return notifier_from_errno(PTR_ERR(new_trustee));
- kthread_bind(new_trustee, cpu);
- /* fall through */
- case CPU_UP_PREPARE:
- BUG_ON(gcwq->first_idle);
- new_worker = create_worker(gcwq, false);
- if (!new_worker) {
- if (new_trustee)
- kthread_stop(new_trustee);
- return NOTIFY_BAD;
- }
- break;
- case CPU_POST_DEAD:
- case CPU_UP_CANCELED:
- case CPU_DOWN_FAILED:
- case CPU_ONLINE:
- break;
- case CPU_DYING:
- /*
- * We access this lockless. We are on the dying CPU
- * and called from stomp machine.
- *
- * Before this, the trustee and all workers except for
- * the ones which are still executing works from
- * before the last CPU down must be on the cpu. After
- * this, they'll all be diasporas.
- */
- gcwq->flags |= GCWQ_DISASSOCIATED;
- default:
- goto out;
- }
-
- /* some are called w/ irq disabled, don't disturb irq status */
- spin_lock_irqsave(&gcwq->lock, flags);
-
- switch (action) {
- case CPU_DOWN_PREPARE:
- /* initialize trustee and tell it to acquire the gcwq */
- BUG_ON(gcwq->trustee || gcwq->trustee_state != TRUSTEE_DONE);
- gcwq->trustee = new_trustee;
- gcwq->trustee_state = TRUSTEE_START;
- wake_up_process(gcwq->trustee);
- wait_trustee_state(gcwq, TRUSTEE_IN_CHARGE);
- /* fall through */
- case CPU_UP_PREPARE:
- BUG_ON(gcwq->first_idle);
- gcwq->first_idle = new_worker;
- break;
+ switch (action) {
+ case CPU_DOWN_PREPARE:
+ flush_gcwq(gcwq);
+ break;
+ }

- case CPU_POST_DEAD:
- gcwq->trustee_state = TRUSTEE_BUTCHER;
- /* fall through */
- case CPU_UP_CANCELED:
- destroy_worker(gcwq->first_idle);
- gcwq->first_idle = NULL;
- break;

- case CPU_DOWN_FAILED:
- case CPU_ONLINE:
- gcwq->flags &= ~GCWQ_DISASSOCIATED;
- if (gcwq->trustee_state != TRUSTEE_DONE) {
- gcwq->trustee_state = TRUSTEE_RELEASE;
- wake_up_process(gcwq->trustee);
- wait_trustee_state(gcwq, TRUSTEE_DONE);
- }
-
- /*
- * Trustee is done and there might be no worker left.
- * Put the first_idle in and request a real manager to
- * take a look.
- */
- spin_unlock_irq(&gcwq->lock);
- kthread_bind(gcwq->first_idle->task, cpu);
- spin_lock_irq(&gcwq->lock);
- gcwq->flags |= GCWQ_MANAGE_WORKERS;
- start_worker(gcwq->first_idle);
- gcwq->first_idle = NULL;
- break;
- }
-
- spin_unlock_irqrestore(&gcwq->lock, flags);
-
-out:
return notifier_from_errno(0);
}

@@ -3745,7 +3491,8 @@ static int __init init_workqueues(void)
unsigned int cpu;
int i;

- cpu_notifier(workqueue_cpu_callback, CPU_PRI_WORKQUEUE);
+ cpu_notifier(workqueue_cpu_up_callback, CPU_PRI_WORKQUEUE_ACTIVE);
+ hotcpu_notifier(workqueue_cpu_down_callback, CPU_PRI_WORKQUEUE_INACTIVE);

/* initialize gcwqs */
for_each_gcwq_cpu(cpu) {
@@ -3768,9 +3515,7 @@ static int __init init_workqueues(void)
(unsigned long)gcwq);

ida_init(&gcwq->worker_ida);
-
- gcwq->trustee_state = TRUSTEE_DONE;
- init_waitqueue_head(&gcwq->trustee_wait);
+ init_waitqueue_head(&gcwq->idle_wait);
}

/* create the initial worker */
Index: linux-2.6/lib/Kconfig.debug
===================================================================
--- linux-2.6.orig/lib/Kconfig.debug
+++ linux-2.6/lib/Kconfig.debug
@@ -62,6 +62,28 @@ config MAGIC_SYSRQ
keys are documented in <file:Documentation/sysrq.txt>. Don't say Y
unless you really know what this hack does.

+config MAGIC_SYSRQ_FORCE_PRINTK
+ bool "Force printk from Magic SysRq"
+ depends on MAGIC_SYSRQ && PREEMPT_RT_FULL
+ default n
+ help
+ Allow the output from Magic SysRq to be output immediately, even if
+ this causes large latencies. This can cause performance problems
+ for real-time processes.
+
+ If PREEMPT_RT_FULL, printk() will not try to acquire the console lock
+ when interrupts or preemption are disabled. If the console lock is
+ not acquired the printk() output will be buffered, but will not be
+ output immediately. Some drivers call into the Magic SysRq code
+ with interrupts or preemption disabled, so the output of Magic SysRq
+ will be buffered instead of printing immediately if this option is
+ not selected.
+
+ Even with this option selected, Magic SysRq output will be delayed
+ if the attempt to acquire the console lock fails.
+
+ Don't say Y unless you really know what this hack does.
+
config STRIP_ASM_SYMS
bool "Strip assembler-generated symbols during link"
default n
Index: linux-2.6/localversion-rt
===================================================================
--- linux-2.6.orig/localversion-rt
+++ linux-2.6/localversion-rt
@@ -1 +1 @@
--rt14
+-rt15
Index: linux-2.6/arch/x86/kernel/apic/apic.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/apic/apic.c
+++ linux-2.6/arch/x86/kernel/apic/apic.c
@@ -856,8 +856,8 @@ void __irq_entry smp_apic_timer_interrup
* Besides, if we don't timer interrupts ignore the global
* interrupt lock, which is the WrongThing (tm) to do.
*/
- exit_idle();
irq_enter();
+ exit_idle();
local_apic_timer_interrupt();
irq_exit();

@@ -1790,8 +1790,8 @@ void smp_spurious_interrupt(struct pt_re
{
u32 v;

- exit_idle();
irq_enter();
+ exit_idle();
/*
* Check if this really is a spurious interrupt and ACK it
* if it is a vectored one. Just in case...
@@ -1827,8 +1827,8 @@ void smp_error_interrupt(struct pt_regs
"Illegal register address", /* APIC Error Bit 7 */
};

- exit_idle();
irq_enter();
+ exit_idle();
/* First tickle the hardware, only then report what went on. -- REW */
v0 = apic_read(APIC_ESR);
apic_write(APIC_ESR, 0);
Index: linux-2.6/arch/x86/kernel/cpu/mcheck/therm_throt.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/cpu/mcheck/therm_throt.c
+++ linux-2.6/arch/x86/kernel/cpu/mcheck/therm_throt.c
@@ -396,8 +396,8 @@ static void (*smp_thermal_vector)(void)

asmlinkage void smp_thermal_interrupt(struct pt_regs *regs)
{
- exit_idle();
irq_enter();
+ exit_idle();
inc_irq_stat(irq_thermal_count);
smp_thermal_vector();
irq_exit();
Index: linux-2.6/arch/x86/kernel/cpu/mcheck/threshold.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/cpu/mcheck/threshold.c
+++ linux-2.6/arch/x86/kernel/cpu/mcheck/threshold.c
@@ -19,8 +19,8 @@ void (*mce_threshold_vector)(void) = def

asmlinkage void smp_threshold_interrupt(void)
{
- exit_idle();
irq_enter();
+ exit_idle();
inc_irq_stat(irq_threshold_count);
mce_threshold_vector();
irq_exit();
Index: linux-2.6/arch/x86/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/irq.c
+++ linux-2.6/arch/x86/kernel/irq.c
@@ -180,8 +180,8 @@ unsigned int __irq_entry do_IRQ(struct p
unsigned vector = ~regs->orig_ax;
unsigned irq;

- exit_idle();
irq_enter();
+ exit_idle();

irq = __this_cpu_read(vector_irq[vector]);

@@ -208,10 +208,10 @@ void smp_x86_platform_ipi(struct pt_regs

ack_APIC_irq();

- exit_idle();
-
irq_enter();

+ exit_idle();
+
inc_irq_stat(x86_platform_ipis);

if (x86_platform_ipi_callback)
Index: linux-2.6/kernel/taskstats.c
===================================================================
--- linux-2.6.orig/kernel/taskstats.c
+++ linux-2.6/kernel/taskstats.c
@@ -657,6 +657,7 @@ static struct genl_ops taskstats_ops = {
.cmd = TASKSTATS_CMD_GET,
.doit = taskstats_user_cmd,
.policy = taskstats_cmd_get_policy,
+ .flags = GENL_ADMIN_PERM,
};

static struct genl_ops cgroupstats_ops = {
Index: linux-2.6/arch/x86/kernel/hpet.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/hpet.c
+++ linux-2.6/arch/x86/kernel/hpet.c
@@ -7,6 +7,7 @@
#include <linux/slab.h>
#include <linux/hpet.h>
#include <linux/init.h>
+#include <linux/dmi.h>
#include <linux/cpu.h>
#include <linux/pm.h>
#include <linux/io.h>
@@ -566,6 +567,29 @@ static void init_one_hpet_msi_clockevent
#define RESERVE_TIMERS 0
#endif

+static int __init dmi_disable_hpet_msi(const struct dmi_system_id *d)
+{
+ hpet_msi_disable = 1;
+}
+
+static struct dmi_system_id __initdata dmi_hpet_table[] = {
+ /*
+ * MSI based per cpu timers lose interrupts when intel_idle()
+ * is enabled - independent of the c-state. With idle=poll the
+ * problem cannot be observed. We have no idea yet, whether
+ * this is a W510 specific issue or a general chipset oddity.
+ */
+ {
+ .callback = dmi_disable_hpet_msi,
+ .ident = "Lenovo W510",
+ .matches = {
+ DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
+ DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad W510"),
+ },
+ },
+ {}
+};
+
static void hpet_msi_capability_lookup(unsigned int start_timer)
{
unsigned int id;
@@ -573,6 +597,8 @@ static void hpet_msi_capability_lookup(u
unsigned int num_timers_used = 0;
int i;

+ dmi_check_system(dmi_hpet_table);
+
if (hpet_msi_disable)
return;

Index: linux-2.6/drivers/watchdog/octeon-wdt-main.c
===================================================================
--- linux-2.6.orig/drivers/watchdog/octeon-wdt-main.c
+++ linux-2.6/drivers/watchdog/octeon-wdt-main.c
@@ -402,7 +402,7 @@ static void octeon_wdt_setup_interrupt(i
irq = OCTEON_IRQ_WDOG0 + core;

if (request_irq(irq, octeon_wdt_poke_irq,
- IRQF_DISABLED, "octeon_wdt", octeon_wdt_poke_irq))
+ IRQF_NO_THREAD, "octeon_wdt", octeon_wdt_poke_irq))
panic("octeon_wdt: Couldn't obtain irq %d", irq);

cpumask_set_cpu(cpu, &irq_enabled_cpus);
Index: linux-2.6/arch/mips/cavium-octeon/smp.c
===================================================================
--- linux-2.6.orig/arch/mips/cavium-octeon/smp.c
+++ linux-2.6/arch/mips/cavium-octeon/smp.c
@@ -207,8 +207,9 @@ void octeon_prepare_cpus(unsigned int ma
* the other bits alone.
*/
cvmx_write_csr(CVMX_CIU_MBOX_CLRX(cvmx_get_core_num()), 0xffff);
- if (request_irq(OCTEON_IRQ_MBOX0, mailbox_interrupt, IRQF_DISABLED,
- "SMP-IPI", mailbox_interrupt)) {
+ if (request_irq(OCTEON_IRQ_MBOX0, mailbox_interrupt,
+ IRQF_PERCPU | IRQF_NO_THREAD, "SMP-IPI",
+ mailbox_interrupt)) {
panic("Cannot request_irq(OCTEON_IRQ_MBOX0)\n");
}
}
Index: linux-2.6/arch/x86/include/asm/irqflags.h
===================================================================
--- linux-2.6.orig/arch/x86/include/asm/irqflags.h
+++ linux-2.6/arch/x86/include/asm/irqflags.h
@@ -60,23 +60,24 @@ static inline void native_halt(void)
#include <asm/paravirt.h>
#else
#ifndef __ASSEMBLY__
+#include <linux/types.h>

-static inline unsigned long arch_local_save_flags(void)
+static inline notrace unsigned long arch_local_save_flags(void)
{
return native_save_fl();
}

-static inline void arch_local_irq_restore(unsigned long flags)
+static inline notrace void arch_local_irq_restore(unsigned long flags)
{
native_restore_fl(flags);
}

-static inline void arch_local_irq_disable(void)
+static inline notrace void arch_local_irq_disable(void)
{
native_irq_disable();
}

-static inline void arch_local_irq_enable(void)
+static inline notrace void arch_local_irq_enable(void)
{
native_irq_enable();
}
@@ -102,7 +103,7 @@ static inline void halt(void)
/*
* For spinlocks, etc:
*/
-static inline unsigned long arch_local_irq_save(void)
+static inline notrace unsigned long arch_local_irq_save(void)
{
unsigned long flags = arch_local_save_flags();
arch_local_irq_disable();
Index: linux-2.6/arch/arm/plat-versatile/platsmp.c
===================================================================
--- linux-2.6.orig/arch/arm/plat-versatile/platsmp.c
+++ linux-2.6/arch/arm/plat-versatile/platsmp.c
@@ -37,7 +37,7 @@ static void __cpuinit write_pen_release(
outer_clean_range(__pa(&pen_release), __pa(&pen_release + 1));
}

-static DEFINE_SPINLOCK(boot_lock);
+static DEFINE_RAW_SPINLOCK(boot_lock);

void __cpuinit platform_secondary_init(unsigned int cpu)
{
@@ -57,8 +57,8 @@ void __cpuinit platform_secondary_init(u
/*
* Synchronise with the boot thread.
*/
- spin_lock(&boot_lock);
- spin_unlock(&boot_lock);
+ raw_spin_lock(&boot_lock);
+ raw_spin_unlock(&boot_lock);
}

int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
@@ -69,7 +69,7 @@ int __cpuinit boot_secondary(unsigned in
* Set synchronisation state between this boot processor
* and the secondary one
*/
- spin_lock(&boot_lock);
+ raw_spin_lock(&boot_lock);

/*
* This is really belt and braces; we hold unintended secondary
@@ -99,7 +99,7 @@ int __cpuinit boot_secondary(unsigned in
* now the secondary core is starting up let it run its
* calibrations, then wait for it to finish
*/
- spin_unlock(&boot_lock);
+ raw_spin_unlock(&boot_lock);

return pen_release != -1 ? -ENOSYS : 0;
}
Index: linux-2.6/arch/arm/mach-exynos4/platsmp.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-exynos4/platsmp.c
+++ linux-2.6/arch/arm/mach-exynos4/platsmp.c
@@ -56,7 +56,7 @@ static void __iomem *scu_base_addr(void)
return (void __iomem *)(S5P_VA_SCU);
}

-static DEFINE_SPINLOCK(boot_lock);
+static DEFINE_RAW_SPINLOCK(boot_lock);

void __cpuinit platform_secondary_init(unsigned int cpu)
{
@@ -76,8 +76,8 @@ void __cpuinit platform_secondary_init(u
/*
* Synchronise with the boot thread.
*/
- spin_lock(&boot_lock);
- spin_unlock(&boot_lock);
+ raw_spin_lock(&boot_lock);
+ raw_spin_unlock(&boot_lock);
}

int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
@@ -88,7 +88,7 @@ int __cpuinit boot_secondary(unsigned in
* Set synchronisation state between this boot processor
* and the secondary one
*/
- spin_lock(&boot_lock);
+ raw_spin_lock(&boot_lock);

/*
* The secondary processor is waiting to be released from
@@ -120,7 +120,7 @@ int __cpuinit boot_secondary(unsigned in
* now the secondary core is starting up let it run its
* calibrations, then wait for it to finish
*/
- spin_unlock(&boot_lock);
+ raw_spin_unlock(&boot_lock);

return pen_release != -1 ? -ENOSYS : 0;
}
Index: linux-2.6/arch/arm/mach-msm/platsmp.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-msm/platsmp.c
+++ linux-2.6/arch/arm/mach-msm/platsmp.c
@@ -38,7 +38,7 @@ extern void msm_secondary_startup(void);
*/
volatile int pen_release = -1;

-static DEFINE_SPINLOCK(boot_lock);
+static DEFINE_RAW_SPINLOCK(boot_lock);

void __cpuinit platform_secondary_init(unsigned int cpu)
{
@@ -62,8 +62,8 @@ void __cpuinit platform_secondary_init(u
/*
* Synchronise with the boot thread.
*/
- spin_lock(&boot_lock);
- spin_unlock(&boot_lock);
+ raw_spin_lock(&boot_lock);
+ raw_spin_unlock(&boot_lock);
}

static __cpuinit void prepare_cold_cpu(unsigned int cpu)
@@ -100,7 +100,7 @@ int __cpuinit boot_secondary(unsigned in
* set synchronisation state between this boot processor
* and the secondary one
*/
- spin_lock(&boot_lock);
+ raw_spin_lock(&boot_lock);

/*
* The secondary processor is waiting to be released from
@@ -134,7 +134,7 @@ int __cpuinit boot_secondary(unsigned in
* now the secondary core is starting up let it run its
* calibrations, then wait for it to finish
*/
- spin_unlock(&boot_lock);
+ raw_spin_unlock(&boot_lock);

return pen_release != -1 ? -ENOSYS : 0;
}
Index: linux-2.6/arch/arm/mach-omap2/omap-smp.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-omap2/omap-smp.c
+++ linux-2.6/arch/arm/mach-omap2/omap-smp.c
@@ -29,7 +29,7 @@
/* SCU base address */
static void __iomem *scu_base;

-static DEFINE_SPINLOCK(boot_lock);
+static DEFINE_RAW_SPINLOCK(boot_lock);

void __cpuinit platform_secondary_init(unsigned int cpu)
{
@@ -43,8 +43,8 @@ void __cpuinit platform_secondary_init(u
/*
* Synchronise with the boot thread.
*/
- spin_lock(&boot_lock);
- spin_unlock(&boot_lock);
+ raw_spin_lock(&boot_lock);
+ raw_spin_unlock(&boot_lock);
}

int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
@@ -53,7 +53,7 @@ int __cpuinit boot_secondary(unsigned in
* Set synchronisation state between this boot processor
* and the secondary one
*/
- spin_lock(&boot_lock);
+ raw_spin_lock(&boot_lock);

/*
* Update the AuxCoreBoot0 with boot state for secondary core.
@@ -70,7 +70,7 @@ int __cpuinit boot_secondary(unsigned in
* Now the secondary core is starting up let it run its
* calibrations, then wait for it to finish
*/
- spin_unlock(&boot_lock);
+ raw_spin_unlock(&boot_lock);

return 0;
}
Index: linux-2.6/arch/arm/mach-tegra/platsmp.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-tegra/platsmp.c
+++ linux-2.6/arch/arm/mach-tegra/platsmp.c
@@ -29,7 +29,7 @@

extern void tegra_secondary_startup(void);

-static DEFINE_SPINLOCK(boot_lock);
+static DEFINE_RAW_SPINLOCK(boot_lock);
static void __iomem *scu_base = IO_ADDRESS(TEGRA_ARM_PERIF_BASE);

#define EVP_CPU_RESET_VECTOR \
@@ -51,8 +51,8 @@ void __cpuinit platform_secondary_init(u
/*
* Synchronise with the boot thread.
*/
- spin_lock(&boot_lock);
- spin_unlock(&boot_lock);
+ raw_spin_lock(&boot_lock);
+ raw_spin_unlock(&boot_lock);
}

int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
@@ -66,7 +66,7 @@ int __cpuinit boot_secondary(unsigned in
* set synchronisation state between this boot processor
* and the secondary one
*/
- spin_lock(&boot_lock);
+ raw_spin_lock(&boot_lock);


/* set the reset vector to point to the secondary_startup routine */
@@ -102,7 +102,7 @@ int __cpuinit boot_secondary(unsigned in
* now the secondary core is starting up let it run its
* calibrations, then wait for it to finish
*/
- spin_unlock(&boot_lock);
+ raw_spin_unlock(&boot_lock);

return 0;
}
Index: linux-2.6/arch/arm/mach-ux500/platsmp.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-ux500/platsmp.c
+++ linux-2.6/arch/arm/mach-ux500/platsmp.c
@@ -57,7 +57,7 @@ static void __iomem *scu_base_addr(void)
return NULL;
}

-static DEFINE_SPINLOCK(boot_lock);
+static DEFINE_RAW_SPINLOCK(boot_lock);

void __cpuinit platform_secondary_init(unsigned int cpu)
{
@@ -77,8 +77,8 @@ void __cpuinit platform_secondary_init(u
/*
* Synchronise with the boot thread.
*/
- spin_lock(&boot_lock);
- spin_unlock(&boot_lock);
+ raw_spin_lock(&boot_lock);
+ raw_spin_unlock(&boot_lock);
}

int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
@@ -89,7 +89,7 @@ int __cpuinit boot_secondary(unsigned in
* set synchronisation state between this boot processor
* and the secondary one
*/
- spin_lock(&boot_lock);
+ raw_spin_lock(&boot_lock);

/*
* The secondary processor is waiting to be released from
@@ -110,7 +110,7 @@ int __cpuinit boot_secondary(unsigned in
* now the secondary core is starting up let it run its
* calibrations, then wait for it to finish
*/
- spin_unlock(&boot_lock);
+ raw_spin_unlock(&boot_lock);

return pen_release != -1 ? -ENOSYS : 0;
}
Index: linux-2.6/include/linux/workqueue.h
===================================================================
--- linux-2.6.orig/include/linux/workqueue.h
+++ linux-2.6/include/linux/workqueue.h
@@ -254,9 +254,10 @@ enum {
WQ_MEM_RECLAIM = 1 << 3, /* may be used for memory reclaim */
WQ_HIGHPRI = 1 << 4, /* high priority */
WQ_CPU_INTENSIVE = 1 << 5, /* cpu instensive workqueue */
+ WQ_NON_AFFINE = 1 << 6, /* free to move works around cpus */

- WQ_DYING = 1 << 6, /* internal: workqueue is dying */
- WQ_RESCUER = 1 << 7, /* internal: workqueue has rescuer */
+ WQ_DYING = 1 << 7, /* internal: workqueue is dying */
+ WQ_RESCUER = 1 << 8, /* internal: workqueue has rescuer */

WQ_MAX_ACTIVE = 512, /* I like 512, better ideas? */
WQ_MAX_UNBOUND_PER_CPU = 4, /* 4 * #cpus for unbound wq */
Index: linux-2.6/drivers/tty/serial/cpm_uart/cpm_uart_core.c
===================================================================
--- linux-2.6.orig/drivers/tty/serial/cpm_uart/cpm_uart_core.c
+++ linux-2.6/drivers/tty/serial/cpm_uart/cpm_uart_core.c
@@ -1225,7 +1225,7 @@ static void cpm_uart_console_write(struc
{
struct uart_cpm_port *pinfo = &cpm_uart_ports[co->index];
unsigned long flags;
- int nolock = oops_in_progress;
+ int nolock = oops_in_progress || sysrq_in_progress;

if (unlikely(nolock)) {
local_irq_save(flags);
Index: linux-2.6/drivers/tty/sysrq.c
===================================================================
--- linux-2.6.orig/drivers/tty/sysrq.c
+++ linux-2.6/drivers/tty/sysrq.c
@@ -492,6 +492,23 @@ static void __sysrq_put_key_op(int key,
sysrq_key_table[i] = op_p;
}

+#ifdef CONFIG_MAGIC_SYSRQ_FORCE_PRINTK
+
+int sysrq_in_progress;
+
+static void set_sysrq_in_progress(int value)
+{
+ sysrq_in_progress = value;
+}
+
+#else
+
+static void set_sysrq_in_progress(int value)
+{
+}
+
+#endif
+
void __handle_sysrq(int key, bool check_mask)
{
struct sysrq_key_op *op_p;
@@ -500,6 +517,9 @@ void __handle_sysrq(int key, bool check_
unsigned long flags;

spin_lock_irqsave(&sysrq_key_table_lock, flags);
+
+ set_sysrq_in_progress(1);
+
/*
* Raise the apparent loglevel to maximum so that the sysrq header
* is shown to provide the user with positive feedback. We do not
@@ -541,6 +561,9 @@ void __handle_sysrq(int key, bool check_
printk("\n");
console_loglevel = orig_log_level;
}
+
+ set_sysrq_in_progress(0);
+
spin_unlock_irqrestore(&sysrq_key_table_lock, flags);
}

Index: linux-2.6/include/linux/sysrq.h
===================================================================
--- linux-2.6.orig/include/linux/sysrq.h
+++ linux-2.6/include/linux/sysrq.h
@@ -38,6 +38,11 @@ struct sysrq_key_op {
int enable_mask;
};

+#ifdef CONFIG_MAGIC_SYSRQ_FORCE_PRINTK
+extern int sysrq_in_progress;
+#else
+#define sysrq_in_progress 0
+#endif
#ifdef CONFIG_MAGIC_SYSRQ

/* Generic SysRq interface -- you may call it from any device driver, supplying
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/